How to Lie Using Statistics

There are three kinds of lies: lies, damned lies, and statistics.  

Benjamin Disraeli


Story by the Numbers

Statistics add credibility to claims, and when utilizing a statistical narrative, you can “prove” anything simply by exploiting a few glitches in thinking. Storytelling is one way. The brain craves a story, and when numbers are manipulated to paint a more convenient picture — three chemicals come into play.


At first, there’s cortisolWebMD refers to cortisol as “nature’s built-in alarm system.” It’s designed to keep you alert when there’s a distressing signal nearby. To manipulate it, you’d require something that catches your attention (like a threat). This is where scary numbers are used on headlines to grab your attention. In advertising, this would be the A, in the famous A.I.D.A. formula: “ATTENTION.” (Not to be confused with the Verdi’s Opera “Aida”).


And then comes the dopamine, which rewards us with great pleasure when we experience the emotionally-packed events of a story unfold. It’s designed to reward motivated behavior. In other words, if you stick to the story and watch it climax, you will feel rewarded. Then, oxytocin allows us to identify who is the protagonist you’re rooting for. Basically, the feeling of empathy for the hero and your mounting desire for him or her to succeed.


“My lab wondered if we could “hack” the oxytocin system to motivate people to engage in cooperative behaviors . . . we tested if narratives shot on video . . . would cause the brain to make oxytocin. By taking blood draws before and after the narrative, we found that character-driven stories do consistently cause oxytocin synthesis. Further, the amount of oxytocin released by the brain predicted how much people were willing to help others; for example, donating money to a charity associated with the narrative.

— Paul J. Zak article “Why Your Brain Loves Good Storytelling
Harvard Business Review


Paul J. Zak, director at the Center for Neuroeconomics Studies at Claremont University, authored an article on the Harvard Business Review website called “Why Your Brain Loves Good Storytelling” which brings up an interesting point. You can steer someone into making a donation by manipulating the narrative. Now, this isn’t supposed to be scary, but makes me wonder what could happen in the hands of the wrong people. Zak continues to comment on how these findings on the neurobiology of storytelling are useful in business settings.



Statistical Narrative


Figures don’t lie; liars figure. – Mark Twain


Statistical narrative means telling an emotionally rich story with numbers — basically bringing quantitative stories to life. Statisticians know that the numbers alone will not necessarily concoct the best argument. Instead, telling a story that lives in the data will yield a better response. “Statistics are fuel for your stories, but they are not stories in themselves,” writes Justin Baker at Borrowing from Justin’s terrific article, he quotes Michael J. Mauboussin from his book “The Success Equation” with the following:


“Once something has occurred and we can put together a story to explain it, it starts to seem like the outcome was predestined.” 

Michael J. Mauboussin in “The Success Equation”


Numbers strengthen arguments; words are too ambiguous. You want your numbers to come from a reliable source. Sometimes, the “data” comes from surveys or polls, which are missing a lot of important factors. For example, it is useful to ask: who asked these questions? Who are these people being surveyed? Did they experience any biases when answering these polls? Otherwise, how useful a poll / survey would be if it’s brimming with biased answers useless to the conclusion.


The intention is not only to discover plain, objective data; but also, to make biases explicit. The goal is to utilize the data to create a fair argument. But, if the goal is to push an idea, the data can be manipulated for that purpose, as well, unfortunately.


Then there’s data storytelling via visualization, such as graphs, infographics, presentations, visualizations, dashboards, etc. Numbers have significant stories embedded in the data that can only be extracted with a narrative.


Three elements come into play: visualizations, data, and narrative. Sometimes they’re used alone, sequentially, or in conjunction. If you’d like to explain something, you simply combine the narrative with the data itself. This way, you can push insights into an audience. When you want to take it a step further, you visualize the data with the narrative.


For example, an infographic can have a combination of said three elements. There’s something about the visuals combined with the data that captures the eye. According to Wikipedia, “infographics can improve cognition by utilizing graphics that enhance the human visual system’s ability to see patterns and trends.”


Your brain is built to accept this information. It’s pleasant and tantalizes the built-in chemicals in your brain that will make it stick (or at least increase the likelihood of “stickability”). Why do you think there are so many infographics shared on social media? Because when people scan Facebook or Twitter on their phone, they’re more likely to stop and consume information from a visual element than just text. If the catchy visuals provide excellent data with an intriguing narrative, the indoctrination is complete.


Let’s look at two infographics. One pro-vaping, and one anti-vaping. Starting with the latter, there’s this one named “The Scary Truth About Teen Vaping.” You can see the data, visuals, and a thread of narrative weaving throughout the infographic. Now, on the other hand, let’s look at the pro-vaping infographic titled “Public Health England Study Says Vaping is 95% Less Harmful than Smoking.” The same principles apply. Each one pushes a specific agenda, fills in only information that supports it, and the data is presented in an eye-catching manner.


Going down the rabbit hole of infographics is time consuming, especially trying to pit anti-something against a pro-something. But if I love motorcycles (especially the Triumph Bonneville) and I see an infographic about the ten best things about owning a Bonnie. My bias as a Bonneville owner is more likely to lead me to click the infographic and read what it has to tell me.


The sad truth is, the brain can be susceptible to deceptive data if its combined with excellent storytelling and eye-catching visuals. This combination can engage the brain long enough for it to believe the data – even if the data is biased and / or incomplete.


“We should all recognize that finding the truth of a situation is an iterative process. It takes time, and trial and error. Our first few attempts at capturing what’s true are almost never right.”

Cliff Kuang “A Case Study in How Infographics Can Bend The Truth



If I asked you to describe your life to me, you would probably tell me anecdotes of your life – possibly as a collection of events. To you, it’s your life, your emotions are linked to each moment; your identity, and your entire psychological make-up stems from these stories.  To me, it’s just a story. The way my brain processes it, it’s not unlike watching a movie. I know it’s a real life story, so there’s definitely more weight to situations and choices, but my brain will fire almost the same as when consuming fiction.


The often called “wonder drug of storytelling” is dopamine. It delivers pleasure to your brain when it follows emotionally potent events of a story. Knowing this, what if you realized that companies and conglomerates use these elements of story to push your buttons to cause you to behave in ways that increase their profit and loss statement? You are well aware of this already, I’m sure. This isn’t a new concept or something corporations are trying to hide anymore.


But the deeper we dig into how we consume information, the more depressing it gets. Statistics have always been a tool of increasing content credibility. Copywriters, journalists, scientists, lots of people use them to increase the trustworthiness of their articles and words — including this very article.


Statistics are weaponized to collect clicks, sales, and shares. For example, I sometimes use studies that reinforce the science behind vaping as an ideal tool for smoking cessation. In my situation, it’s because that was my own personal path. That is confirmation bias. It’s not always a bad thing. The problem arises when it is used as a sleight of hand. When they know the bias and use the statistic to steer it towards an agenda.



If I had to write a hypothetical article called “Ten Reasons to Love Vaping”, I would pick the ten best benefits. Why would I mention the bad and the ugly? That would be an example of playing on your confirmation bias. I use it, and I fall for it. Our human nature is wildly malleable and gullible, even when we think we’re well equipped against trickery.


In a video by YouTube channel MajorPrep titled “This is How Easy It Is to Lie with Statistics, the host tells a story of a Target statistician who developed a mathematical system with the purpose of determining the patterns of women expecting a baby (whether they wanted to divulge that information or not).


The statistician discovered a common pattern which consisted of expecting mothers buying products like lotions, specific vitamins, etc. After a while, the patterns would allow them to see who’s likely to be expecting, and could even determine the timeline down to which trimester they were in.


With an eye close to this data, Target would then create coupons selling items like baby powders, strollers, etc., and send to these women. To avoid suspicion, they would also sprinkle other items in the coupons so it would appear random instead of targeted. The story continues with the case study involving an angry man who screamed at one store’s general manager because he believed they were trying to encourage his teenage daughter to have a baby by sending her these coupons. When the manager followed up a few days later to apologize for the situation, the angry man apologized in turn and explained that his daughter, was, in fact, pregnant. The statistics knew better than the girl’s own father.


With this data, Target would send coupons at specific times when certain items were highly likely to be required for the baby. Let’s say, hypothetically, if during 2019 only one person out of a thousand developed “popcorn lung” from vaping. Then, by the end of 2020 two people developed the disease. The statistic here is a 100% increase. You can imagine the headline already, can’t you?



On the flip side, for those on the side of vaping, could use this statistic to illustrate how low the number of popcorn lung cases truly exist. With only 2 cases in the entirety of the market segment of e-cigarette users, it’s nothing to panic about. We will use the figure that paints the most convenient picture.


Correlation vs. Causation


58.6% of all statistics are made up on the spot

– unknown


When I was a child, I remember my mother telling me I should stop wearing hats because I’d go bald. Her reasoning: “All these bald men we know wear hats all the time. Therefore, if you wear hats, you’ll go bald, too.”  The truth is, they probably wear hats to cover up their cue balls. In this particular situation of ignorant mum and son, it seems likely that it is an assumption of correlation. We see bald men wearing hats, so we assume hats caused the baldness.


When looking at the science behind this problem, it appears that wearing tight headgear could cause stress on the hair follicle, which could lead to the hair falling off. It’s something called traction alopecia. So, mum was right. Well, not exactly, because hats don’t create the tension necessary to cause traction alopecia or other forms of baldness. I’m sure she meant well.


This same situation applies to the common “violent videogames create violent children” argument. Will violent videogames turn an otherwise meek child into a violent kid, or is it the innate violent tendencies of some children to gravitate towards violent video games in the first place?


It’s the classic Post Hoc fallacy: If A-Event leads to B-Event, then A-Event causes B-Event. Example, “I walked in the snow without my shoes on and now I have a cold. Therefore, the snow gave me a cold.” The cold temperature (amongst other factors) affect your immune system, making you more susceptible to the virus. It’s not the snow itself “giving” you the cold. The virus was already in your system; the low temperature simply lowered your defenses enough for the virus to take hold give you the sniffles.


Oftentimes, when we see two things together, we assume one leads to the other, or one causes another. Correlation vs. causation is not that simple. Sometimes, you can experience two seemingly correlated issues that have an outside causation. This is something called third-cause fallacy.


Let’s say you have a group of obese people and you’re trying to determine a pattern. You discover all of them are wealthy with no evidence to the contrary (for the time being). You build the assumption: wealthy people are obese. What if, in this hypothetical scenario, the group is obese because of an outside factor? Maybe this group shares a common third factor like divorced parents, friendship issues, weather, etc., and the stress of this leads to overeating (and they happen to have the means to really indulge)?


Whatever the case may be, sometimes, hiding the outside causality (third-cause fallacy) can help someone make an argument for or against something. If they have an anti-vaping agenda, they could utilize the exploding vapes statistics illustrating the equation: A leads to B (Vaping leads to explosions and injury), when there’s always a different cause of user ignorance and misinformation that leads to the purchasing and mishandling of a dangerous device.


“The following is a simple example: ‘The rooster crows immediately before sunrise, therefore the rooster causes the sun to rise’.  This conclusion is false, not just we happen to know that it is factually incorrect, but because the argument is fallacious.”

– The Logical Place, “Confusing correlation with causation.”




“Fallacies, fallacies. Run for you, dead for me, all your lies won't set you free from fallacies, fallacies. “
– Jesse Pinkman’s song “Fallacies” from Breaking Bad


A fallacy is a failure in reasoning, or flawed thinking, and can take many, many forms. Sometimes an argument doesn’t have many legs to stand on, and plenty of imagery and statistical trickery is used to conceal the gap. When encountering shady statistics, it’s good to look for fallacies. As you interrogate a premise, you can interrogate it further with critical thinking and break it apart with questions and an objective eye.


 The Writing Center at the University of North Carolina at Chapel Hill defines fallacies as “defects that weaken arguments. By learning to look for them in your own and others’ writing, you can strengthen your ability to evaluate arguments you make, read, and hear.” The author continues to describe how fallacious arguments can be persuasive to a casual reader in advertising and other popular mediums.


An argument is only strong by the number of fallacies it lacks. When you detect a weak argument, you will see assumptions being thrown without evidence, or based on incomplete data.


As stated in the U.S. Represented video titled “logical fallacies,” and, some of the most common are:


Ad Hominem: attacking a person instead of the issue at hand, i.e. name calling, insults, etc.

Hasty Generalization: drawing a conclusion based on few data.

False Dichotomy: when thinking there are only two choices when, in fact, there are alternative choices.

Begging the Question: Using an opinion as fact, and the assumption of the initial point.

Appeal to Authority: When someone accepts a truth blindly because they admire the person or outlet who said it.

Red Herring: When someone uses unrelated facts and information to steer away from the argument.

Slippery Slope: The assumption of a small action leading to catastrophic outcomes.

And many more…


Right off the bat, we can remember some of these fallacies on things we’ve read at some point. For example, the argument that vaping causes seizures. You can find the slipper slope fallacy, “vaping leads to seizures” which is not as black and white as that. The science behind seizures is more complicated and the small action cannot lead to the seizure directly. There is more nuance to discuss before coming up with a conclusion.


We all know friends, coworkers, and even ourselves; accepting opinions and data as FACTS right away, simply because we like and trust the person or outlet who said it. I am a victim of using, and falling for, most of these from time to time. Hell, even in this writing.  


Confirmation Bias

There are up to 53 cognitive biases, and one of the most common is the confirmation bias. As we know, a bias works as a shortcut of sorts by taking the shortest way into processing certain information in the favor for or against something. In this case, we only accept information that confirms our existing beliefs. It’s convenient when we believe something, and if we read an article or data supporting what we strongly believe, we are confirming our bias. It’s like a craving for the justification of an existing belief.


According to the highly-popular YouTube channel Practical Psychology, a confirmation bias “leads people to seek information that confirms their pre-existing beliefs or opinions.” This bias is weaponized in data storytelling and statistics in the form of testimonials, case studies, and whatnot.


Other forms include articles and media focusing only on the benefits, and the positive side. The unbiased solution here would be to seek both the negative and positive from the pool of information and seek the clearest view on the issue. From this collection of untainted, unbiased data, you can draw your conclusion.


“Confirmation biases impact how we gather information, but they also influence how we interpret and recall information. For example, people who support or oppose a particular issue will not only seek information to support it, they will also interpret news stories in a way that upholds their existing ideas. They will also remember details in a way that reinforces these attitudes.”

Kendra Cherry, “Confirmation Bias: We interpret facts to confirm beliefs


Interpreting new information through a biased lens can be overwhelmingly limiting. Imagine all the new ideas and great information you leave in your mental spam folder.


Here is how I am a slave to confirmation bias

When writing articles about vaping, I always think of vaping as an NRT (Nicotine Replacement Therapy) because that was my path. Having stopped smoking cigarettes, I still vape, except now it’s only nicotine free e-liquids.


My confirmation bias leads me to seek case studies, articles, and science which supports this. I want to confirm my beliefs, over and over. I do, however, understand the negative sides of vaping such as harms of nicotine and other chemicals. But when looking at the big picture, going from smoking cigarettes to nicotine-free vaping, it’s a whole new world. That is what I seek to confirm, which is not useful to everyone.


Let’s use data storytelling, infographics and confirmation bias for the following tweet: